Influence of dissolved gas concentration on the lifetime of surface bubbles in volatile liquids

Bubbles at the air-liquid interface are important for many natural and industrial processes. Factors influencing the lifetime of such surface bubbles have been investigated extensively, yet the impact of dissolved gas concentration remains unexplored. Here we investigate how the lifetime of surface bubbles in volatile liquids depends on the dissolved gas concentration. The bubble lifetime is found to decrease with the dissolved gas concentration. Larger microbubbles at increased gas concentration are found to trigger bubble bursting at earlier times. Combined with the thinning rate of the bubble cap thickness, a scaling law of the bubble lifetime is developed. Our findings may provide new insight on bubble and foam stability.

💡 Research Summary

The paper investigates how dissolved gas concentration influences the lifetime of surface bubbles in volatile liquids, using isopropanol as a model fluid. By preparing liquids with varying levels of dissolved oxygen through vacuum degassing (undersaturation) and pressurization (oversaturation), the authors quantify gas oversaturation as ζ = (c – c_sat)/c_sat, where c is the measured dissolved oxygen concentration and c_sat = 8.3 mg L⁻¹ at 23 °C and 1 atm. Surface bubbles are generated by injecting a fixed volume of air (90 µL, with additional tests at 25 µL and 55 µL) through a 0.5 mm needle into the liquid, forming a hemispherical cap that rests on the liquid surface. The experiments are conducted in petri dishes (polyethylene) and quartz containers, with liquid depths of 10, 20, and 40 mm, to test the robustness of the observed trends.

Bubble lifetimes (t_l) are recorded with a standard camera at 60 fps, while high‑speed imaging (3 000–25 000 fps) captures the rupture dynamics. The liquid film thickness of the bubble cap is measured using the Taylor‑Culick relation h = 2σ/(ρv²), where σ and ρ are the surface tension (20.4 mN m⁻¹) and density (0.781 g cm⁻³) of isopropanol, and v is the hole‑retraction speed observed in the high‑speed footage. Initial film thickness h_i, measured shortly after bubble formation, remains essentially constant at ≈0.93 µm across all ζ values, indicating that dissolved gas does not affect the initial cap geometry.

The key finding is that t_l decreases monotonically with increasing ζ. The decline is steep in the undersaturated regime (ζ < 0) and becomes more gradual as the liquid approaches saturation and then oversaturation. Simultaneously, the film thickness at rupture h_r increases with ζ, reaching up to ≈0.8 µm, still below h_i, implying that bubbles burst earlier when the film is thicker.

High‑speed observations reveal that rupture is almost always triggered by a small spherical object rising through the film. The authors identify this object as a microbubble rather than a solid impurity, based on its spherical shape and disappearance after rupture. Statistical analysis shows that 92.8 % of recorded bursting events involve such a particle, and the true probability is likely higher because some particles are hidden from view.

Microbubble diameters (d) are measured in the bulk just before they intersect the cap. The data show a near‑linear increase of d with ζ, consistent with diffusion‑controlled growth described by the Epstein‑Plesset model (d ∝ c). The authors argue that when d exceeds the instantaneous film thickness h_r by a factor k > 1 (i.e., d/h_r ≈ k), the microbubble can locally thin the film enough to nucleate a hole, leading to rupture.

Combining these observations, the authors propose a scaling law for bubble lifetime: \